1. Lecture 4. Niching and Speciation (1)
학습목표
진화로 얻어진 해의 다양성을 확보하기 위한 대표적인 방법에 대해 이해한다.

2. Outline Review of the last lecture Different types of co-evolution
Different applications
Learning (e.g., games, classification, …)
Optimization (e.g., sorting networks)
Design (e.g., floor plan design, …)
A motivating example of niching: co-evolution in classification tasks
Why niching
Different niching techniques
Relationship between niching and speciation
Summary

3. A Motivating Example (1)
Task:
Design a machine learning system to classify two classes. A training set is given, consisting of 97 examples of class A and 97 examples of class B.
Assume that you are going to evolve this machine learning system (which can be a rule-based system or a neural network). How do you develop such an evolutionary system?
Representation:
Search (Genetic) operators:
Fitness evaluation:

4. Why Niching
In evolutionary computation, niching refers to the formation of groups of individuals in a population
Individuals within a group are similar to each other
Individuals from different groups are dissimilar to each other
Niching helps to explore and maintain diversity
Niching helps in machine learning, e.g., classification
Niching helps multi-objective function optimization
Niching helps simulation of complex and adaptive systems
Niching helps cooperative co-evolution
Niching is fun!

5. Different Niching Techniques
Can be divided roughly into two major categories
Sharing, also known as fitness sharing
Crowding
Other niching methods include sequential niching and parallel hillclimbing

6. Fitness Sharing: Introduction
Fitness sharing transforms the raw fitness of an individual into shared fitness
It assumes that there is only limited and fixed “resource” available at each niche. Individuals in a niche must share them
Sharing is best explained from a multimodal function optimization perspective
raw
fitness
individual
How can we locate multiple peaks in one evolutionary process?